bigyabai

17 ms·

421.
by bigyabai 3mo ago
> How are Chinese companies financing their buildouts of AI infrastructure? The same way that OpenAI and Anthropic will dig themselves out of the hole; By pawning off the training costs onto taxpaying citizens. > Well you don't n
422.
by bigyabai 3mo ago
> You can't simultaneously project a hypothetical Chinese capability that it hasn't proven it can do (develop the leading models) China does lead the frontier. If you're a cybersecurity researcher, a biologist or work in a
423.
by bigyabai 3mo ago
I consider it more humane than paying humans to do the same work a-la Mechanical Turk, so in a relative sense yes.
424.
by bigyabai 3mo ago
> they don't understand the notion of truth so they can't do verifiable work like math proofs. Nobody that understands automated proof checking was claiming that.
425.
by bigyabai 3mo ago
> Opus 4.6 remains the best model because of this. Huh? Are you not trying other open-weight models that stream thinking traces? The new DeepSeek-V4-Flash-0731 should clobber Opus 4.6 in a lot of tasks. Kimi K3 and GLM 5.2 feel like they
426.
by bigyabai 3mo ago
> It's not wish casting whatever that means. Wishcasting is describing a hypothetical. Your statement was about what you think America can do, but not about what it has proven to do already. It is wishcasting by any definition you
427.
by bigyabai 3mo ago
> but even with those we only went through about 5% of the stockpile. Source? In terms of VLS-packed high end interceptors, I'd imagine they're much deeper into the magazine than that. Quad-packed point defense missiles might b
428.
by bigyabai 3mo ago
> They'll offer far superior models at a premium price > US companies will also release open source models and lower-cost options This is wishcasting. People are raising the alarm because models like GLM-5.2 and Kimi K3 have alr
429.
by bigyabai 3mo ago
> I [...] imagine for some employers that makes you a red flag Maybe. It also advertises a moral flexibility that is high in-demand here in America. Clearly OpenAI has no scruples entertaining a turncoat of this scale, I can believe that
430.
by bigyabai 3mo ago
The DMA enriches PWA software. There's nothing in it poised to kill the tech.
431.
by bigyabai 3mo ago
In fairness, that's an incredibly glaring MDM problem that was always going to blow up like a landmine if Apple didn't address it. I'd expect the macOS MDM policy to cover this if Apple was actually concerned about IP leakage
432.
Native Apps Should Be Avoided Whenever Possible(nooneshappy.com)
42 points by bigyabai 3mo ago|30 comments
433.
by bigyabai 3mo ago
This "Fields medalist" work is automated proof-checking. Turning the question around, then - what does it say about you to worship pushbutton smoke-testing? I don't what your life is like, but I don't salivate with dis
434.
by bigyabai 3mo ago
So abundant that you won't link anything? Send a peer-reviewed paper then, let's settle this.
435.
by bigyabai 3mo ago
This would be a better comment if your first post on this website wasn't literally a sarcastic NIMBY driveby: https://news.ycombinator.com/item?id=48261228 The way you hurl stones from a glass house it kinda just sound
436.
by bigyabai 3mo ago
> The scale of the data center buildout makes no sense. It's a fully commoditized service. Selling compute is an incredibly lucrative business and investors want in, it's just as simple as that.
437.
by bigyabai 3mo ago
One is a derivative work.
438.
by bigyabai 3mo ago
> We are now at the point where RSI is feasible What can be asserted without evidence can also be dismissed without evidence. - Hitchen's Razor
439.
by bigyabai 3mo ago
> I think you are unable to cope with AI encroaching your life and you are hiding away from it. That would be the case if you were doing anything impressive or noteworthy with it. Can you show me your AI magnum opus, or are you also st
440.
by bigyabai 3mo ago
No.
441.
by bigyabai 3mo ago
Pack-in slopware and it's consequences have been a disaster for the desktop operating system duopoly.
442.
by bigyabai 3mo ago
It's perfectly coherent. For instance; mRNA vaccines changed your life, why are you ignoring it to bloviate about LLMs? How is that coherent? AI has had a much smaller impact on humanity, relatively speaking. Suffice it to say that H
443.
by bigyabai 3mo ago
Yes but I don't consider most of them to be a problem.
444.
by bigyabai 3mo ago
Just my anecdata, I've seen dozens of CS students that own a Steam Deck in the past few years. It's a perfectly suitable alternative to owning a gaming laptop for many students.
445.
by bigyabai 3mo ago
You can put SteamOS on any hardware. Your old gaming desktop, a mini PC, weird crypto mining hardware - it's all a testement to how flexible Proton is. This project is pretty involved, but by no means is it the easiest alternative.
446.
by bigyabai 3mo ago
It's been great. I hope they keep importing them, Qwen and Z.ai have done far more for humanity with that compute than Anthropic or OpenAI ever did.
447.
by bigyabai 3mo ago
Sure, on the same grounds that you can own any gun. When you lug it into public and threaten ordinary people, you're liable to be followed, questioned or arrested.
448.
by bigyabai 3mo ago
I mean, you don't put the explosives in until the end. Until then it's an "ironic hobbyist UAV" or whatever you want to call it.
449.
by bigyabai 3mo ago
> If you call this stance “anti-AI,” again, I encourage you to read the declaration: it specifically promotes responsible development and use of AI to improve human well-being. But "responsible" is an arbitrary definition that
450.
by bigyabai 3mo ago
If my PC or phone did that, I would immediately begin shopping for a replacement. I'm not three.

More ›